SAN FRANCISCO — April 10, 2026
Executive Summary
Kastle AI has announced a direct integration with Intercontinental Exchange’s (ICE) mortgage servicing system, MSP®, enabling mortgage servicers to deploy AI agents within their existing servicing infrastructure. The integration allows servicers using MSP to securely access servicing data and automate workflows through Kastle’s platform without requiring changes to their core systems. Through this integration, Kastle’s AI agent Avery can retrieve loan data, execute servicing actions, and support borrower interactions directly within MSP. The announcement reflects Kastle’s continued expansion across the mortgage servicing ecosystem, providing AI-driven capabilities for borrower engagement, collections, and operational processes while maintaining alignment with established systems of record.
Announcement Overview
Kastle AI has introduced a new integration with Intercontinental Exchange’s (ICE) MSP® mortgage servicing system, designed to extend AI capabilities into existing mortgage servicing environments. The integration enables mortgage servicers currently operating on MSP to deploy Kastle’s AI agents directly within their servicing workflows, allowing secure interaction with loan data and servicing functions.
Through this integration, Kastle’s platform connects with MSP to facilitate both data retrieval and execution of servicing actions. The system enables AI agents to operate within the established servicing infrastructure, ensuring continuity with existing processes while introducing automation across high-volume servicing activities. The integration also supports real-time access to loan-level information, allowing AI-driven workflows to be executed with accuracy and operational alignment.
The announcement positions the integration as a mechanism for enabling mortgage servicers to incorporate AI technologies without requiring custom development or infrastructure changes. By leveraging MSP as the system of record, servicers can deploy AI capabilities while maintaining consistency with their current servicing environments.
Key Announcement Details
Core Announcement Identification
- Announcement Title: Kastle Announces Direct Integration with the ICE MSP Mortgage Servicing System to Expand AI Adoption in Mortgage Servicing
- Announcement Date: April 10, 2026
- Announcement Location: San Francisco, California
- Announcing Organization: Kastle AI
- Integration Counterparty: Intercontinental Exchange (ICE)
- Integrated System: MSP® mortgage servicing system
- Announcement Category: AI integration within mortgage servicing infrastructure
- Primary Objective: Expansion of AI adoption in mortgage servicing environments
Integration Framework and Functional Access
- Integration Type: Direct integration with ICE MSP system
- System Role of MSP: Mortgage servicing system used by servicers as system of record
- Integration Capability: Enables deployment of Kastle AI agents within MSP
- Data Access Mode: Secure access to servicing data through integration
- Execution Capability: AI agents can execute servicing actions within MSP
- Retrieval Capability: AI agents can retrieve loan-level data from MSP
- Activation Requirement: Integration operates once authorized by servicers
- Infrastructure Requirement: No need for custom infrastructure build-out
- Architecture Impact: No changes required to existing core servicing systems
- Operational Environment: AI agents operate directly within MSP environment
- Connectivity Type: Direct connectivity to MSP without intermediary system replacement
Avery AI Agent – Core Functional Scope
- Agent Name: Avery
- Agent Type: Servicing AI agent
- Operational Role: Executes servicing tasks and interacts with borrowers and systems
- System Interaction: Direct interaction with MSP servicing system
- Borrower Interaction: Direct engagement with borrowers
Loan Data and Servicing Execution
- Loan Information Retrieval: Retrieves loan data directly from MSP
- Servicing Action Execution: Executes servicing functions within MSP
Payment Processing
- Borrower Payment Processing: Processes borrower payments through MSP
Contact Establishment
- QRPC Creation: Establishes Quality Right Party Contacts (QRPCs)
Task Management
- Task Creation: Creates servicing tasks directly within MSP
- Workflow Participation: Integrates into MSP workflow processes
Documentation and System Recording
- Loan Activity Documentation: Records servicing activity within MSP workstations
- System Screen Logging: Documents activity within servicing screens
Real-Time Data Utilization
- Real-Time Data Retrieval: Retrieves loan data in real time
- Workflow Execution Support: Uses real-time data for servicing workflows
Platform Deployment and Market Position
- Platform Category: AI agent platform for mortgage servicing
- Deployment Status: Among the most deployed AI agent platforms in mortgage servicing
- Geographic Market: United States
- Customer Segment: Leading mortgage servicers
Servicing Workflow Automation Coverage
- Borrower Engagement Automation: Automates borrower interaction workflows
- Collections Workflow Automation: Supports collections processes
- Operational Quality Control Automation: Automates servicing QA/QC workflows
Agent Sentinel – Quality Assurance and Compliance Automation
- Agent Name: Agent Sentinel
- Functional Area: Quality assurance and compliance monitoring
Call Quality Assurance
- Call QA Automation: Automates review of borrower calls
- Coverage Level: Provides full coverage of call QA processes
Document Quality Control
- Document QC Automation: Automates servicing document review
- Coverage Level: Provides full coverage of document QC
Operational Impact
- Manual Review Reduction: Reduces dependence on large manual QA/QC teams
- Process Coverage: Enables comprehensive QA/QC across servicing operations
Real-Time AI Co-Pilot Functionality
- Feature Type: Real-time AI co-pilot
- User Group: Human servicing agents
Functional Capabilities
- Conversation Assistance: Assists agents during borrower interactions
- Operational Context Support: Provides real-time guidance during servicing workflows
Integration Benefits and Operational Enablement
- AI Adoption Enablement: Allows servicers to adopt AI agents within MSP
- Infrastructure Simplicity: Eliminates need for custom infrastructure
- Secure Data Access: Provides secure access to servicing data
- Workflow Execution Enablement: Enables execution of servicing workflows through Kastle platform
Servicing Functions Enabled
- Borrower Engagement: Automation of borrower communication
- Payment Processing: Automation of payment workflows
- Collections Outreach: Automation of collections activities
- Operational QA: Automation of quality assurance processes
System Integrity and Data Governance
- System of Record Preservation: MSP remains the system of record
- Data Source Integrity: Loan data retrieved directly from MSP
- Execution Integrity: All servicing actions executed within MSP
- Documentation Integrity: Activities recorded within MSP systems
- Data Duplication Avoidance: No external duplication of servicing data
Strategic Purpose and Adoption Objective
- Adoption Goal: Accelerate AI adoption among MSP-based servicers
- Deployment Flexibility: Enables activation without system modification
- Operational Expansion: Supports expansion of automated servicing workflows
- Technology Alignment: Aligns AI capabilities with existing servicing systems
Leadership Statement – Kastle CEO
- Executive Name: Rishi Choudhary
- Title: Chief Executive Officer
- Company: Kastle
Statement Highlights
- MSP Role Recognition: Identified as core system of record for servicers
- Integration Outcome: Enables deployment of AI agents like Avery within MSP
- Operational Continuity: Allows servicers to use AI within existing systems
ICE Positioning and Disclaimer
- Ownership Statement: ICE does not own Kastle platform
- Control Statement: ICE does not control industry participants
- Endorsement Statement: ICE does not endorse products or services
Responsibility Allocation
- Vendor Selection Responsibility: Loan originators and servicers must vet providers
- Contracting Responsibility: Servicers responsible for contracting with providers
Kastle Platform – Functional Overview
- Company Name: Kastle
- Platform Type: AI platform for consumer lending and banking
- Technology Focus: Domain-specific AI agents
Core Functional Areas
- Customer Service Automation: Automates customer service operations
- Originations Support: Supports loan origination workflows
- Back-Office Automation: Automates back-office functions
Operational Benefits
- Efficiency Improvement: Enhances operational efficiency
- Cost Reduction: Supports cost reduction initiatives
- Customer Experience Enhancement: Improves borrower experience
Company Background and Investment
- Founding Support: Backed by Y Combinator
- Additional Backing: Commerce Ventures
Headquarters
- Location: San Francisco, California
Integration with ICE MSP Mortgage Servicing System
The integration connects Kastle’s AI platform directly with MSP®, the mortgage servicing system operated by Intercontinental Exchange (ICE). MSP serves as a system of record for many mortgage servicers, supporting core servicing operations and data management across loan portfolios.
Through this integration, Kastle enables AI agents to interact with MSP in a secure and structured manner. The connection allows for both retrieval of loan information and execution of servicing actions, enabling AI-driven workflows to function within the same system used for day-to-day servicing operations.
The integration is structured to ensure that servicers can maintain their existing system architecture while introducing AI capabilities. This approach eliminates the need for custom-built infrastructure, allowing organizations to adopt automation technologies within their current operational frameworks.
By aligning with MSP, Kastle’s platform integrates into a widely used servicing environment, enabling consistent data access and operational execution across servicing processes.
AI Agent Avery: Core Servicing Capabilities Within MSP
Avery, Kastle’s servicing AI agent, plays a central role in the integration. The agent is designed to interact directly with both borrowers and servicing systems, enabling the execution of common mortgage servicing tasks within MSP.
Using the ICE integration, Avery can autonomously retrieve loan information and execute servicing actions, supporting a range of operational workflows. The agent operates within the servicing environment, accessing relevant data and performing actions in alignment with system processes.
Key Functional Capabilities of Avery Within MSP
- Processing borrower payments directly within the servicing system
- Establishing QRPCs (Quality Right Party Contacts)
- Creating tasks within MSP workflows and workstations
- Documenting loan activity across appropriate servicing screens
- Retrieving loan data in real time to support workflow execution
These capabilities enable Avery to function as an operational agent within the servicing lifecycle, handling both borrower-facing and backend processes. The integration ensures that all actions are performed within MSP, maintaining system-of-record consistency while enabling automation.
Avery’s ability to interact with both borrowers and servicing systems allows for continuous workflow execution, reducing reliance on manual processes for high-frequency servicing tasks.
Broader AI Agent Platform and Operational Coverage
Beyond Avery, Kastle’s AI platform includes additional agents designed to support various aspects of mortgage servicing operations. These agents extend automation capabilities across quality assurance, compliance monitoring, and real-time support for servicing teams.
Agent Sentinel: Quality Assurance and Compliance Monitoring
Agent Sentinel is designed to automate quality assurance (QA) and compliance monitoring across borrower communications and servicing documentation. The agent enables comprehensive coverage of call QA and document quality control processes.
- Automates review of borrower calls for compliance and quality
- Monitors servicing documentation for accuracy and adherence
- Provides full coverage of QA and QC processes traditionally handled manually
- Reduces reliance on large manual review teams
By automating QA and QC processes, Agent Sentinel supports operational consistency and compliance across servicing activities.
Real-Time AI Co-Pilot for Servicing Agents
Kastle also provides a real-time AI co-pilot that assists human servicing agents during borrower interactions.
- Supports agents during live borrower conversations
- Provides contextual assistance based on servicing data
- Enhances decision-making during servicing workflows
The co-pilot functionality enables human agents to operate with additional support, improving efficiency while maintaining alignment with servicing requirements.
Together, these AI agents form a comprehensive platform designed to automate and support multiple layers of mortgage servicing operations.
Deployment Without Custom Infrastructure Requirements
A key feature of the integration is the ability for mortgage servicers to deploy Kastle’s AI agents without building custom infrastructure. The integration allows Kastle to connect directly with MSP once authorization is granted, enabling secure access to loan data and immediate execution of servicing workflows.
Operational Advantages of the Integration
- No requirement for custom system development
- Direct access to servicing data within MSP
- Immediate activation of AI-driven workflows upon authorization
- Alignment with existing servicing processes and architecture
This deployment model allows servicers to introduce AI capabilities without modifying their core servicing systems. By leveraging the integration, organizations can implement automation within their current operational environment.
The ability to deploy AI without infrastructure changes supports faster adoption and reduces the complexity associated with integrating new technologies into existing systems.
Enabling AI Adoption Across Mortgage Servicing Workflows
The integration is designed to support the adoption of AI across multiple servicing functions, enabling mortgage servicers to automate key workflows while maintaining operational consistency.
Servicing Workflows Supported by the Integration
- Borrower engagement and communication
- Payment processing workflows
- Collections outreach and follow-ups
- Operational quality assurance and compliance monitoring
By enabling automation across these areas, the integration allows servicers to extend AI capabilities throughout the servicing lifecycle. The use of MSP as the system of record ensures that all activities remain aligned with existing data structures and operational processes.
The integration also supports real-time workflow execution, allowing AI agents to operate with current loan data and perform actions within the servicing system. This ensures that automation is integrated into ongoing operations rather than functioning as a separate layer.
Servicers can leverage Kastle’s platform to expand their operational capabilities while maintaining consistency with established servicing systems.
Maintaining System of Record Integrity Within MSP
A central aspect of the integration is the preservation of system of record integrity within MSP. All AI-driven actions performed by Kastle’s platform occur within MSP, ensuring that servicing data remains consistent and centralized.
Key Aspects of System Integrity
- All servicing actions executed within MSP
- Loan data retrieved directly from the system of record
- Documentation recorded within appropriate servicing screens
- No duplication of data outside the core system
This approach ensures that AI-driven workflows operate within the same environment used for traditional servicing processes. By maintaining system integrity, servicers can adopt automation without disrupting existing data management practices.
The integration aligns AI capabilities with established servicing systems, ensuring continuity across operations.
Leadership Statement
Rishi Choudhary, CEO of Kastle, stated:
“MSP is the system of record that many mortgage servicers depend on every day. This integration allows servicers to deploy AI agents like Avery using the same system they already rely on for servicing operations.”
This statement highlights the integration’s focus on enabling AI adoption within existing systems, emphasizing compatibility with established servicing infrastructure.
ICE Clarification on Integration and Provider Independence
The announcement includes a clarification regarding the role of Intercontinental Exchange (ICE) in relation to the integration and participating service providers.
Key Clarifications
- ICE does not own or control Kastle’s platform
- ICE does not endorse specific industry participants
- ICE does not endorse any product or service provided
- Loan originators and servicers are responsible for vendor selection and contracting
This clarification establishes that the integration operates within the MSP ecosystem without implying endorsement or control by ICE. Mortgage servicers retain responsibility for evaluating and selecting service providers based on their requirements.
The statement ensures transparency regarding the relationship between the MSP platform and third-party providers integrating with the system.
About Kastle
Kastle is an AI platform developed for consumer lending and banking applications. The platform delivers domain-specific AI agents designed to automate customer service, loan originations, and back-office workflows.
Core Platform Capabilities
- Automation of customer service interactions
- Support for loan origination processes
- Execution of back-office operational workflows
- Delivery of compliant AI solutions tailored to financial services
Kastle partners with financial institutions across the United States, supporting operational efficiency and workflow automation through its AI-driven platform.
Company Background
- Backed by Y Combinator and Commerce Ventures
- Headquartered in San Francisco
- Focused on AI applications within consumer lending and banking
The company’s platform is designed to support financial institutions in managing servicing operations while integrating automation into existing workflows.
Media Contact
For additional information, visit kastle.ai.
Source Attribution
Source: Company announcement
